Coupling the P-Graph Framework with Life Cycle Assessment: Synthesis of Biofuels and Biochemicals Co-Production

While prospective life-cycle assessment (LCA) represents an important step towards understanding the environmental impacts of emerging technologies, its evaluation of processes is inherently retrospective – pre-determined by... [ view full abstract ]

While prospective life-cycle assessment (LCA) represents an important step towards understanding the environmental impacts of emerging technologies, its evaluation of processes is inherently retrospective – pre-determined by the underlying environmental performance of the proposed process alternatives. Conversely, process synthesis approaches, such as the P-graph method within Process Network Synthesis (PNS), offer systematic frameworks to design process alternatives, but traditionally ignore life-cycle environmental implications. In coupling prospective LCA with PNS, we may perform rigorous chemical process design through the lens of various life-cycle metrics, rather than assessing processes after completion of heuristic-driven design. In constructing a preliminary case study, we observe emerging multistage torrefaction and fast pyrolysis systems, which have shown promise in lab-scale and process modeling studies. Such systems could produce drop-in replacement biofuels with diesel-range hydrocarbon profiles, while achieving over a 75% reduction in life-cycle GHG emissions relative to petroleum diesel and a promising median Energy Return On Investment (EROI) ranging from 1.14 to 3.17 MJ-Fuel/MJ-Primary Fossil Energy. Expanding upon these multistage systems by allowing for co-production of biofuels and biochemicals – shown to potentially further improve life-cycle environmental performance – PNS can be used to formulate optimal designs for a host of life-cycle environmental metrics. 

This work proposes novel coupling of PNS and LCA methodologies for simultaneous design and evaluation of emerging drop-in replacement biofuel technologies. The full design space of multistage torrefaction of short rotation poplar biomass to bio-oil followed by catalytic upgrading to biofuels and biochemicals is investigated. In formulating optimal process designs for an array of life-cycle objectives, the P-graph framework is applied to construct the maximal superstructure of co-production strategies as a directed bipartite graph, with unit operation and material type nodes. Multi-objective optimization is performed on this graph to assess trade-offs between objective functions for each life-cycle metric, resulting in solution fronts containing values for each metric. Life-cycle objectives are sourced from the aforementioned prospective LCA and calculated via its process parameters and life-cycle inventory data, including Ecoinvent  and USLCI.  The array of life-cycle objectives over which multi-objective optimization is conducted includes: (1) life-cycle GHG minimization, (2) fuel EROI maximization, and (3) effective carbon yield maximization. By way of simultaneous PNS and prospective LCA of these emerging energy and chemical production technologies, design for and evaluation of their environmental performance is made explicit.

Utilizing PNS provides distinct advantages over traditional heuristic-based approaches for process design, especially where the P-graph encompasses known, promising designs alongside unknown pathways. Various sequences and groupings of material streams and unit operations, along with conditions within them, are produced based on the specified importance of a given objective. Trade-offs between process complexity and life-cycle environmental performance are observed and reported for different priorities of stated objectives. Specific advantages of the proposed approach over heuristic-based process design and process LCA will be discussed to justify its application in future studies of emerging technologies.